Article Text


Local intra-arterial fibrinolysis for acute occlusion of the central retinal artery: a meta-analysis of the published data
  1. S Beatty,
  2. K G Au Eong
  1. Academic Department of Ophthalmology, Manchester Royal Eye Hospital
  1. S Beatty, Medical Retina Unit, Department of Clinical Ophthalmology, Moorfields Eye Hospital, City Road, London EC1V 2PDstephen{at}


BACKGROUND/AIM Central retinal artery occlusion (CRAO) is typically associated with a poor visual outcome. Several favourable reports of local intra-arterial fibrinolysis (LIF), which involves the superselective administration of a thrombolytic agent directly into the ophthalmic artery, have appeared in the recent literature. The aim of this study was to critically appraise these studies in a collective fashion.

METHODS A meta-analysis was performed of all the published literature germane to LIF in cases of CRAO.

RESULTS Of the 16 studies identified, all were retrospective and non-randomised. After correction for data duplication, the results of LIF in 100 patients can be reported. A final acuity of 6/6 or better was seen in 14% of patients following LIF, and a visual result of 6/12 or better was seen in 27% of subjects. A poor final acuity of 3/60 or worse was seen in 60.6% of eyes treated with local intra-arterial fibrinolysis. These results compare favourably with conventional forms of therapy. Potentially serious complications were seen in four patients, but no patient suffered a permanent neurological deficit.

CONCLUSION The results of this study suggest that there may be a marginal visual benefit associated with LIF compared with conventional management of CRAO. However, the methodology of the cited studies was often unsatisfactory, and a randomised controlled trial of LIF in cases of CRAO is justified. Outside of a randomised clinical trial, the use of superselective fibrinolytic therapy for CRAO cannot be recommended on the basis of current evidence.

  • central retinal artery occlusion
  • fibrinolysis
  • streptokinase
  • thrombolysis
  • tissue plasminogen activator
  • urokinase

Statistics from

Central retinal artery occlusion (CRAO) is associated with a poor visual prognosis, and aggressive management with ocular massage, anterior chamber paracentesis, and carbogen therapy does not appear to improve the outcome.1 2 These considerations have motivated the use of catheter administered, local intra-arterial fibrinolysis (LIF) for CRAO, of which there are several encouraging reports in the literature. We have performed a meta-analysis of the published data in order to investigate whether we should be offering LIF to our patients presenting with acute occlusive events of the retinal arterial circulation.


Sixteen reports of intra-arterial fibrinolytic therapy for CRAO were identified in the published literature,3-18 (Table1), and covert duplicate use of data was found in three of these papers3 6 8 and acknowledged duplication in a further two papers,7 10 a problem commonly encountered in meta-analyses.19 Therefore, only the most recent and complete series reported by these investigators were included in the analysis. Although we do comment on case reports and small series (less than five subjects), the meta-analysis is confined to the five studies involving seven or more subjects.

Table 1

Studies and review articles relating to fibrinolytic therapy in central retinal artery occlusion

The technique of LIF varied between studies, but in all cases the fibrinolytic agent was injected by hand through a coaxial catheter system placed in the ophthalmic artery. The dose was typically determined by the need to avoid reaching therapeutic systemic levels, and by clinical response as evaluated by repeated ophthalmoscopy, confrontational perimetry, and acuity testing during the procedure. LIF was always followed by heparinisation for a period of 2–3 days.


The visual acuities at presentation were statistically similar for the studies that satisfied the inclusion criteria (χ219.89; p=0.07), and the original data were therefore combined and analysed as a single file comprising 100 subjects with CRAO treated with local intra-arterial fibrinolytic therapy. Ages ranged from 19 to 87 years, with a mean (SD) of 61.1 (13.9), and the male:female ratio was 71:29. The mean delay between the onset of symptoms and the administration of fibrinolytic therapy was 11.6 (8.7) hours (range 3–60 hours). Fibrinolytic agents used included urokinase (42; 42%) and recombinant tissue plasminogen activator (rt-PA) (58; 58%). All patients had a Snellen visual acuity of 6/60 or worse at presentation, with the exception of two who had acuities of 6/18 and 6/9

Final visual acuities ranged from perception of light (PL) to 6/6. Of the 100 subjects, 14 achieved a final acuity of 6/6 or better, 27 achieved a final acuity of 6/12 or better, and 37 achieved a final vision of 6/60 or better (Fig 1). These results represent a mean improvement in acuity of three (3.05) Snellen lines. Changes in acuity were unrelated to the delay between onset of symptoms and administration of the fibrinolytic agent (simple regression analysis:r=0.135; p=0.27), or to the type of agent used (ANOVA: F=2.54; p=0.11). However, changes in Snellen vision did differ between groups of subjects categorised according to presenting acuities (ANOVA: F=3.43; p=0.02). In general, poor initial acuity was a poor prognostic indicator (Fig 1). Seven of 48 eyes (14.6%) with PL or no perception of light (NPL) at presentation achieved a final Snellen acuity of 6/36 or better, whereas nine of 32 eyes (28.1%) with presenting acuities of hand movements (HM) or counting fingers (CF) achieved this level of vision. There is no dramatic change in the results if we include case reports and small series.

Figure 1

Scattergram illustrating the relation between visual outcome and presenting acuity following local intra-arterial fibrinolysis for central retinal artery occlusion. In order to generate a scattergram, categories of Snellen acuity were ascribed an integer score ranging from 1 (NPL) to 13 (6/6 or better). Random noise of not more than plus or minus 0.5 was then superimposed onto the results to ensure that a maximum number of data points were visually represented. NPL = no perception of light, PL = perception of light, HM = hand movements, CF = counting fingers.

Of the 100 patients reported, and accounting for duplication of data, complications were seen in six (6%). Complications included haemorrhage at puncture track of femoral artery (one),6hemiplegia with recovery (three),3 18 and hypertensive crisis with recovery (one).18 All cases of hemiplegia were treated with immediate fibrinolytic therapy following angiographic identification of the occluded vessel, thus averting a permanent neurological deficit. Of note, no cases of cerebral or retinal bleeding were reported.


The aim of LIF in cases of CRAO is to restore retinal blood supply by dissolving the occluding thromboembolus, and rt-PA has become the agent of choice for this procedure because of its shorter half life and minimal effects on physiological clotting, and because its local administration reduces the risk of systemic side effects.20 The rationale of LIF rests on the assumption that the damage caused by retinal ischaemia is reversible. A retinal tolerance time of 105 minutes has been demonstrated in monkeys following total occlusion of the central retinal artery,21and up to 4 hours in the presence of a moderate amount of residual retinal circulation.22 In humans, however, significant visual improvement occurring 48 hours or more following conventional treatment for CRAO has been reported in eight of 32 cases reviewed by Augsburger and Magargal, although the visual prognosis was noted to worsen with increasing duration of visual symptoms.23Another concern rests on the fact that only 15.5% of retinal emboli are composed of platelet fibrin, the remainder being made up of cholesterol (74.5%) and calcium (10.5%).24 As fibrinolytic agents have no effect on cholesterol or calcium, it has been postulated that LIF may be appropriate for only a minority of patients.22 However, it should be noted that the most cases of CRAO are thrombotic and not embolic, and that the stasis induced by non-platelet fibrin emboli can result in the formation of secondary thrombi which contribute to the arterial occlusion.25

Of the 16 studies reporting on the use of LIF for CRAO, all are retrospective and non-randomised, and only two compared the visual outcome of LIF with the natural course of disease in control subjects.3 5 The results of our meta-analysis indicate that intra-arterial fibrinolytic therapy in cases of acute CRAO is associated with a marginally better visual outcome than conservative forms of management. For example, a final acuity of 6/6 or better was seen in 14% of patients following LIF compared with 9% for the natural course of CRAO,22 and a visual result of 6/12 or better was seen in 27% of subjects compared with 18% to 21% reported for a variety of conventional treatments.23 26 In one of the two controlled studies, a marked improvement in acuity (equals five Snellen lines) was seen in five of 15 patients treated with intra-arterial urokinase (33%) but in none of the 17 subjects treated with traditional methods.5 However, these findings should be interpreted with full appreciation of the possibility that publication bias may have limited the reporting of unfavourable results.

The questionable rationale of LIF for CRAO, and the inconclusive results of this meta-analysis, are reflected in a lively debate on the subject in a recent issue ofOphthalmology.22 Hayreh, in response to the favourable report of Richard et al of LIF in cases of acute occlusion of the central retinal artery,18 expresses several concerns. In addition to the issues of retinal tolerance time, study design, and embolus composition, Hayreh argues that fundus fluorescein angiography is a “fundamental requirement for any study dealing with a retinal circulatory disorder” in order to establish the amount of residual retinal blood flow before fibrinolysis; the presenting acuity may be deceptively poor because testing is done under less than ideal conditions in an emergency situation, and the patient has not yet learnt to fixate eccentrically; the lack of a relation between visual outcome and time lapse between the occlusive event and administration of fibrinolytic therapy is inconsistent with published scientific studies.27 In brief, Hayreh believes the results reported by Richard et al simply represent the natural history of CRAO.22

Nevertheless, CRAO remains a visually debilitating condition for which we have no effective treatment. If LIF was to result in even a marginally better visual outcome than conventional management, it would be a welcome advance. It is likely that a substantial proportion of patients would consent to undergo this emergency procedure as a recent survey has shown that 37% of binocular adults with CRAO would risk a cerebrovascular accident or death in order to triple their chances of recovering a visual acuity of 6/36 or better in one eye, and this rose to 80% for monocular subjects.28

We need and should support a randomised controlled trial (RCT) of LIF for CRAO if the debate regarding the risks and benefits of this procedure is to be brought to a satisfactory conclusion. In our view, there is currently insufficient evidence to justify LIF outside of a RCT.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.